Obstructive sleep apnea (OSA) is a multifactorial disorder with probably four main causes or physiologic traits: 1) an anatomically small, or collapsible, upper airway, 2) an unstable ventilatory control system, 3) a low respiratory arousal threshold from sleep, and 4) a poor upper airway muscle response during sleep. The broad term objective of this research is to better understand how these traits interact to produce OSA in individual patients, and then to use this information to design new treatments. Specifically, this grant aims to validate a novel technique for measuring and modeling the traits causing OSA (Aim 1), and then determine how effective non-continuous positive airway pressure therapies are at manipulating the traits (Aim 2). To achieve these objectives, the within-night and between-night repeatability of the measurements and model will be tested, along with the validity of the ventilatory stability and arousal threshold metrics. Note: the upper airway anatomy/collapsibility measurement has been previously validated and will not be repeated. Also, the upper airway (muscle) response measurement will not be validated because the methodology is straightforward. The ventilatory stability metric (loop gain), will be validated by administering hypoxic gas to individuals and comparing the hypoxic loop gain to the normoxic loop gain (hypoxia raises the loop gain). As there is not a gold standard for measuring loop gain, the new method is being validated by determining if it can detect a directional change in loop gain. The new arousal threshold measurement will be validated by comparing it to the arousal threshold determined from esophageal pressure monitoring. In Aim 2, the effect of several interventions on each trait will be measured. The interventions that will be tested include upper airway surgery and oral appliances (to manipulate pharyngeal collapsibility), acetazolamide and supplemental oxygen (to manipulate the control of breathing), and eszopiclone (to manipulate the arousal threshold). Interventions to manipulate the upper airway muscle response will not be tested since currently there are not good candidate drugs for doing this. The studies proposed will not only improve our understanding of OSA pathophysiology, but could realistically lead to new therapeutic approaches.